Computers conventionally use rotating magnetic media for mass storage of documents, data, programs and information. Though widely used and commonly accepted, such hard disk drives suffer from a variety of deficiencies. Because of the rotation of the disk, there is an inherent latency in extracting information from a hard disk drive.
Other problems are especially dramatic in portable computers. In particular, hard disks are unable to withstand many of the kinds of physical shock that a portable computer will likely sustain. Further, the motor for rotating the disk consumes significant amounts of power decreasing the battery life for portable computers.
Solid state memory is an ideal choice for replacing a hard disk drive for mass storage because it can resolve the problems cited above. Potential solutions have been proposed for replacing a hard disk drive with a semiconductor memory. For such a system to be truly useful, the memory must be nonvolatile and alterable. The inventors have determined that FLASH memory is preferred for such a replacement.
FLASH memory is a transistor memory cell which is programmable through hot electron, source injection, or tunneling, and erasable through Fowler-Nordheim tunneling. The programming and erasing of such a memory cell requires current to pass through the dielectric surrounding a floating gate electrode. Because of this, such types of memory have a finite number of erase-write cycles. Eventually, the dielectric deteriorates. Manufacturers of FLASH cell devices specify the limit for the number of erase-write cycles between 100,000 and 1,000,000.
One requirement for a semiconductor mass storage device to be successful is that its use in lieu of a rotating media hard disk mass storage device be transparent to the designer and the user of a system using such a device. In other words, the designer or user of a computer incorporating such a semiconductor mass storage device could simply remove the hard disk and replace it with a semiconductor mass storage device. All presently available commercial software should operate on a system employing such a semiconductor mass storage device without the necessity of any modification.
SunDisk proposed an architecture for a semiconductor mass storage using FLASH memory at the Silicon Valley PC Design Conference on Jul. 9, 1991. That mass storage system included read-write block sizes of 512 Bytes to conform with commercial hard disk sector sizes.
Earlier designs incorporated erase-before-write architectures. In this process, in order to update a file on the media, if the physical location on the media was previously programmed, it has to be erased before the new data can be reprogrammed.
This process would have a major deterioration on overall system throughput. When a host writes a new data file to the storage media, it provides a logical block address to the peripheral storage device associated with this data file. The storage device then translates this given logical block address to an actual physical block address on the media and performs the write operation. In magnetic hard disk drives, the new data can be written over the previous old data with no modification to the media. Therefore, once the physical block address is calculated from the given logical block address by the controller, it will simply write the data file into that location. In solid state storage, if the location associated with the calculated, physical block address was previously programmed, before this block can be reprogrammed with the new data, it has to be erased. In one previous art, in erase-before-write architecture where the correlation between logical block address given by the host is one to one mapping with physical block address on the media. This method has many deficiencies. First, it introduces a delay in performance due to the erase operation before reprogramming the altered information. In solid state flash, erase is a very slow process.
Secondly, hard disk users typically store two types of information, one is rarely modified and another which is frequently changed. For example, a commercial spread sheet or word processing software program stored on a user's system are rarely, if ever, changed. However, the spread sheet data files or word processing documents are frequently changed. Thus, different sectors of a hard disk typically have dramatically different usage in terms of the number of times the information stored thereon is changed. While this disparity has no impact on a hard disk because of its insensitivity to data changes, in a FLASH memory device, this variance can cause sections of the mass storage to wear out and be unusable significantly sooner than other sections of the mass storage.
In another architecture, the inventors previously proposed a solution to store a table correlating the logical block address to the physical block address. The inventions relating to that solution are disclosed in U.S. Pat. No. 5,388,083, issued on Feb. 7, 1995. U.S. Pat. No. 5,479,638 issued on Dec. 26, 1995. Those applications are incorporated herein by reference.
The inventors' previous solution discloses two primary algorithms and an associated hardware architecture for a semiconductor mass storage device. It will be understood that “data file” in this patent document refers to any computer file including commercial software, a user program, word processing software document, spread sheet file and the like. The first algorithm in the previous solution provides means for avoiding an erase operation when writing a modified data file back onto the mass storage device. Instead, no erase is performed and the modified data file is written onto an empty portion of the mass storage.
The semiconductor mass storage architecture has blocks sized to conform with commercial hard disk sector sizes. The blocks are individually erasable. In one embodiment, the semiconductor mass storage can be substituted for a rotating hard disk with no impact to the user, so that such a substitution will be transparent. Means are provided for avoiding the erase-before-write cycle each time information stored in the mass storage is changed.
According to the first algorithm, erase cycles are avoided by programming an altered data file into an empty block. This would ordinarily not be possible when using conventional mass storage because the central processor and commercial software available in conventional computer systems are not configured to track continually changing physical locations of data files. The previous solution includes a programmable map to maintain a correlation between the logical address and the physical address of the updated information files.
All the flags, and the table correlating the logical block address to the physical block address are maintained within an array of CAM cells. The use of the CAM cells provides very rapid determination of the physical address desired within the mass storage, generally within one or two clock cycles. Unfortunately, as is well known, CAM cells require multiple transistors, typically six. Accordingly, an integrated circuit built for a particular size memory using CAM storage for the tables and flags will need to be significantly larger than a circuit using other means for just storing the memory.
The inventors proposed another solution to this problem which is disclosed in U.S. Pat. No. 5,485,595, issued on Jan. 16, 1996. That application is incorporated herein by reference.
This additional previous solution invented by these same inventors is also for a nonvolatile memory storage device. The device is also configured to avoid having to perform an erase-before-write each time a data file is changed by keeping a correlation between logical block address and physical block address in a volatile space management RAM. Further, this invention avoids the overhead associated with CAM cell approaches which require additional circuitry.
Like the solutions disclosed above by these same inventors, the device includes circuitry for performing the two primary algorithms and an associated hardware architecture for a semiconductor mass storage device. In addition, the CAM cell is avoided in this previous solution by using RAM cells.
Reading is performed in this previous solution by providing the logical block address to the memory storage. The system sequentially compares the stored logical block addresses until it finds a match. That data file is then coupled to the digital system. Accordingly, the performance offered by this solution suffers because potentially all of the memory locations must be searched and compared to the desired logical block address before the physical location of the desired information can be determined.
What is needed is a semiconductor hard disk architecture which provides rapid access to stored data without the excessive overhead of CAM cell storage.